Measuring Creativity: a Case Study in the UK's Designer Fashion Sector

Cultural Trends Vol. 14(2), No. 54, June 2005, pp. 157–183

Measuring Creativity: A Case Study in the UK’s Designer Fashion Sector Anne Creigh-Tyte

Since before the 1997 General Election, New Labour has repeatedly emphasized the importance of the creative industries in underpinning the national economy and as an engine of economic growth. The Creative Industries Task Force Mapping Documents of 1998 and 2001 sought to define and quantify in broad terms economic activity across 13 distinct creative industries. More detailed estimates have been published by the Depart- ment for Culture, Media and Sport (DCMS) in successive annual Creative Industries Economic Estimates. An assessment is provided of the way in which the creative industries have been measured using the Office of National Statistics’ Standard Industrial and Standard Occupational classifications (SIC and SOC). These classifications have themselves been revised since the early 1990s, and further revisions are in prospect from 2007. In this analysis particular focus is given to the ‘Designer Fashion’ sector, illustrated by a number of tables and data analyses. These actual and proposed revisions have helped in documenting the rapidly emerging creative industries, which have reportedly grown at two to three times the rate of the UK economy as a whole. However, as the Regional Culture Data Framework published in 2002 records, serious problems remain in providing valid assessments of the creative industries sectors from ‘official’ published sources, even for the UK as a whole, let alone at the regional level emphasized by the Regional Culture Data Framework’s regional spon- sors. In any case, often the ‘scaling factors’ applied to official SIC codes to define creative industries appear arbitrary. Many of the Regional Culture Data Framework’s recommendations, notably the adoption of a more comprehensive ‘supply-chain’ approach to documenting the cultural sector, make further demands upon the existing official structural classifications and the data bases underpinning them. Even where all elements in the ‘supply chain’ are well documented, there are still questions about the validity of this approach. For example, should wholesale and retail distribution of creative industry products be regarded as part of the ‘Cultural Cycle’?

Correspondence to: Dr Anne Creigh-Tyte, Faculty Director of Design Research and the Centre for Fashion Indus- try Research, Graduate and Research Centre, Faculty of Art, Design and Architecture, Kingston University, Knights Park, Kingston Upon Thames, Surrey KT1 2QJ, UK. Email: [email protected]

ISSN 0954-8963 (print)/ISSN 1469-3690 (online) # 2005 Taylor & Francis DOI: 10.1080/09548960500292383 158 A. Creigh-Tyte In conclusion, it is suggested that the ‘ofﬁcial’ data has marked limitations in documenting the creative industries and does not realistically or adequately capture the more interesting and dynamic elements of an industry like ‘Designer Fashion’. This is disappointing in a context where central government has placed increasing emphasis upon evidence-based policy to support the development of the creative industries, and where the British ‘Designer Fashion’ sector has lamented the lack of central support in comparison with the French or Italian industries. It is suggested that a more customized approach to collecting data about the creative industries is needed if the results are to usefully inform the further development and proﬁle of these sectors.

Keywords: Cultural economy; Creative industries; ‘Designer Fashion’; Ofﬁcial statistics; Culture Cycle; Specialist studies

Introduction The incoming Labour Government of 1997 laid particular emphasis on developing the creative industries as an engine of economic growth for the UK. At the same time the new government emphasized the role of evidence-based policy in assessing ‘what works’. Within these contexts this article reviews the attempts made to date to quantify the high-media-profile ‘Designer Fashion’ sector, where there are no directly applicable codes within the Standard Industrial Classification (SIC), which is the official statistical classification used by central government. Beginning with the Creative Industries Task Force’s Mapping Documents, the Department for Culture, Media and Sport (DCMS) has sought to apply the official statistics codes, but serious problems remain. Proposals within the Regional Cultural Data Framework and Departmental Evidence Toolkit to adopt a wider ‘Cultural Cycle’ spreading from creation through to education/understanding in a supply-chain approach across the DCMS’s sectors have only increased the challenges faced by those seeking to document sectors such as ‘Designer Fashion’. Even before the 1997 election, increasing emphasis was placed on the potential of the cultural economy at national and regional levels. The incoming government also placed evidence and research at the heart of its policy agenda, which led to successive attempts to ‘map’ the creative industry sectors. For some creative industries, notably ‘Designer Fashion’, direct SIC and SOC official statistical categories are not available or are inadequate. The predominant method is application of ‘scaling’ functions to orthodox SIC industry codes. Derivation of these ‘scaling’ factors is obscure and their magnitude differs between sources. Alternative approaches have been developed, notably the six-phase ‘Culture Cycle’ model and the introduction of more detailed SIC codes which allow ‘scaling’ to be super- seded. Unfortunately, for many parts of the ‘Culture Cycle’ no appropriate SIC codes are proposed and as yet there are no data available at the most applicable detailed SIC level. For ‘Designer Fashion’ in particular, these limitations in official data are critical, since the various customized studies produced during the 1990s and even the most Cultural Trends 159 recent studies covering 2001 at national and regional levels (North West and Yorkshire and Humberside) all have distinct limitations in terms of transparency and survey response rates.

The Vision The Potential of the ‘Cultural Economy’ The new Labour government of 1997 noted in its pre-election strategy document that:

the arts and cultural industries are among the most proﬁtable sectors of the British economy, directly employing an estimated 648,000 people. The cultural industries for which the current Heritage Department has responsibility have a turnover of more £50 billion a year, almost seven per cent of GDP. (Labour Party, 1997, p. 7) Various commentators stressed that one of the most important parts of the so-called new economy comprised a group of industries identiﬁed as suppliers of cultural products (Scott, 2000). Given the importance of this emerging ‘cultural economy’, it is not surprising that Labour’s strategy document inter alia pledged to

re-form the Department to play a major part in the economic regeneration of our country, working with the cultural industries, local government and the private sector to create wealth and employment and that

the huge importance and value of British architecture and design will be a funda- mental consideration in all government policy. (Labour Party, 1997 p. 7, author’s emphasis) At the end of Labour’s strategy document sat the famous mandala produced by Gorham Associates which encapsulated ‘cultural industries’ as a ‘creative base’ (writing, performance, composing, painting, design), around which the successive rings of ‘live performance’ (e.g. exhibitions) and their distribution and delivery (e.g. broadcasting), ‘associated activities’ (e.g. advertising) and ‘tools of the trade’ (e.g. exhibitions display equipment) revolved. (See Labour Party, 1997, p. 25.)

Evidence-Based Policy Once in office, Labour placed increasing emphasis upon the role of evidence in policy making across all Cabinet Departments. Thus the Modernising Government White Paper (Cabinet Office, 1999), emphasized that government should regard policy-making activity as a continuous learning process, not as a series of one-off initiatives. This improved use of evidence and research should include: . utilizing evidence to inform the development and reform of policies and services; . research aimed at the understanding of broad policy contexts and complex policy areas; . the contribution of forecasting to policy implementation. 160 A. Creigh-Tyte The commitment was given more focus in the Performance and Innovation Unit’s Adding It Up report (Cabinet Office Performance and Innovation Unit (COPIU), 2000) on improving analysis and modelling in central government. Among 41 wide-ranging conclusions, Adding It Up emphasized the need for improved micro- economic modelling as well as more and better data. Government departments, including the DCMS and its sponsored bodies, were expected to provide an evidence base for their policies and the role of evidence was seen as central to policy development. Evidence-based policy was seen as a key in determining ‘what works’ and hence where resources should be focused. (See Davies, Nutley & Smith, 2000, for an overview.) However, it is important to remember that in culture or elsewhere, evidence is itself the end result of a complex analytical process beginning with the collection of raw data (see Table 1). Moreover, the policy development and implementation process involves a series of stages and levels as well as the need to monitor progress, evaluate outcomes and feed back conclusions into future development (see Table 2). Researchers and evidence-based policy practitioners from a wider range of disciplines are inevitably faced by serious time constraints, since the policy cycle typically revolves more quickly than the research cycle (see Hanberger, 2001). The feedback loop is a core, if often in practice neglected, element in the UK Treasury’s standard approach on economic appraisal and evaluation for central government as set out in the Treasury’s latest Green Book (HM Treasury, 2003). Evidence-based policy has been most thoroughly applied in medicine, but even there only about one-third of practice is supported by rigorous random controlled trial studies. It is also important to remember that medical literature is less fragmen- ted, more comprehensively covered by bibliographical data bases, better indexed and subject to a far greater degree of ‘professional consensus’ and ‘received wisdom’ than the social sciences. The social sciences encompass a wide range of professional literature, including ‘grey literature’ reports commissioned or produced by public, private and voluntary bodies. Not surprisingly, criteria for evaluating evidential quality remain elusive in the social sciences. The Campbell Collaboration’s objective is to improve the evidence base for policy in the social sciences by reviewing studies in particular themes. The Campbell Collaboration was inaugurated in 2000 when the Cochrane Collaboration, which has an explicit medical focus, handed over to the Campbell Collaboration a

Table 1 Getting to the Evidence

Raw Data whether counted or measured once codiﬁed, checked and systematically set out/tabulated become Information which in turn when focused on a particular set of hypotheses/policies yields Analysis which when weighed, reviewed, assessed and subject to professional judgement and scrutiny becomes Evidence on the impact of a particular hypothesis, policy/programme/or projects

Source: Creigh-Tyte and Mundy (2003) Cultural Trends 161 Table 2 The Policy Context

Policies are the intervention equivalent of theories predicting behaviour implemented through more detailed Programmes of activities or particular one-off Projects Both programmes and projects need to be subjected to ex-ante expected cost/beneﬁt analyses before implementation. Such appraisal needs to be followed up by Monitoring using mostly administrative data during the implementation stages and full-scale Evaluation of ex-post impact, effectiveness and actual achieved cost/beneﬁt assessment so as to provide lessons and allow Feedback into improved design of future policies/programmes and projects

Source: Creigh-Tyte and Mundy (2003) registry of over 10,000 trials in education, social work/welfare and criminal justice. See Campbell (1974) and Cochrane (1972) for an insight into the development of these data bases. Within the social sciences there are various views of the evidence-based policy model, as illustrated in Table 3. However, government-funded or -inspired evidence- based policy work in the social sciences seems much more likely to fall within models (ii)–(iv) in Table 3, hopefully following the ‘problem-solving’ and ‘interactive’ approaches (i.e. (ii) and (iii)). Every academic research study begins with a review of previous literature, designed at least in part to highlight the contribution of the new research. However, the systematic review as developed initially in the medical context by the Cochrane Collab- oration differs signiﬁcantly from the traditional ‘literature review’ in terms of focus on a key question, and assessment and evaluation of the evidence and its quality, including the methods used.

Table 3 Models of the Evidence-Based Policy Process

Model applicability Role of researcher

(i) Knowledge driven High-technology projects Researcher ‘on top’ (ii) Problem solving Research is shaped by policy Researcher ‘on tap’ priorities (iii) Interactive Mutually inﬂuential Some researchers research and policy inﬂuential, e.g. think ‘communities’ tanks (iv) ‘Realist’ political/tactical Research commissioned Researcher establishes a model and/or used to support point policy position adopted (v) Enlightenment model Serves policy agenda Researcher inspired by urge indirectly by illuminating to understand/explain the landscape for decision stands at a distance makers from immediate policy concerns—provides ‘evidence-informed’ basis for policy

Source: Adapted from Young, Ashby, Boaz and Grayson (2002), in Creigh-Tyte, Gillham and Mundy (2004) 162 A. Creigh-Tyte In practice, rigorous policy research of any type, let alone to rigorous systematic review standards, is scarce in the cultural and creative ﬁelds. Cultural sector researchers will not be surprised to note that a comparative assessment of UK systematic review policy literature produced for the fourth Campbell Collaboration Colloquium characterized the method and debate in terms of a ‘multi-disciplinary area spanning impacts of creativity on non-cultural policy areas covering research using a range of methods. Few experimentation exercises with control groups’ (Creigh-Tyte, Gillham & Mundy, 2004, p. 4). In terms of the nature of the evidence, the same authors concluded that there were ‘strong project appraisal elements but relatively under-developed evaluation work. Numerous superﬁcially relevant studies but few with robust and transparent method. Heavy reliance on case studies but even these are often poorly documented’ (Creigh-Tyte et al., 2004, p. 4).

‘Mapping’ the ‘Vision’ After the 1997 election, a new Creative Industries Task Force (CITF) was established to bring together ministers across the government and a range of industry advisors to consider how government and industry could work together to create an environment for sustainable growth across the creative industries. The CITF defined the creative industries as: ‘... those activities which have their origin in individual creativity, skill and talent and which have a potential for wealth and job creation through the generation and exploitation of intellectual property’ (DCMS & CITF, 1998, p. 003). In practice these activities were taken to include the following key sectors: ‘advertising, architecture, the art and antique market, crafts, design, “Designer Fashion”, film, interactive leisure software, music, the performing arts, publishing, software and television and radio’ (DCMS & CITF, 1998, p. 003). The CITF had five specific objectives:

one to map activity in the UK creative industries and assess its economic value in all parts of the UK, two to promote a wider appreciation of that value, three to consider how current government policies influence activity in the creative industries and what further steps government might take to promote such activity in different parts of the UK, four to look for threats to the continuing growth of the creative industries within the UK and opportunities for increasing the wealth-creating potential of these industries, and five to make recommendations to government, industry and others. (DCMS & CITF, 1998, p. 003) Thus the CITF’s primary objective was to map the creative industry sectors. This was undertaken mainly by Strategy Spectrum Consultants and by City University’s Depart- ment of Arts Policy and Management in the case of the first and second Mapping Documents respectively. Cultural Trends 163 The Statistical Framework Any attempt to apply the evidence-based policy ‘vision’ to the UK’s cultural and creative sectors must, at least in the first resort, rely on the cross-cutting statistical framework applied to all economic activities by central government statisticians. The basis for classifying business establishments, the UK’s SIC was introduced in 1948. It was revised in 1958, 1968, 1980, 1992 and 1997. Such periodic revision is needed over time as new products and industries emerge and there are shifts in emphasis within existing industries. The 1997 SIC revisions were not major in scope but, instead, largely reflected user demand for a number of additional sub-classes. The most recent 2003 publication (SIC, 2003) represents a further limited revision reflecting additional sub-class detail and the revisions to the European Union (EU)’s statistical classification of economic activities of January 2003 (Nomenclature geńere´rale des activite´sećonomiques dans les Communaute´s Europeénnes, hereafter, NACE 1.1). The SIC classifies statistical units according to the category ‘...that best describes their activity, taking into account not only the output structures but also the input structure, including the production process’ (Office for National Statistics (ONS), 2003, p. 11). Ideally, the allocation is by share in value added, but where this is not available sub- stitutes may include gross output or value of sales as alternative output bases, or wages and salaries, or employment as input bases. The SIC allocates units successively on a tree-and-branch basis to sections, divisions, groups, classes and finally sub-classes respectively at the 1-, 2-, 3-, 4- and lowest 5-digit level. Thus two or more 5-digit- level codes aggregate to a 4-digit code, whereas two or more 4-digit code classes can be aggregated to the 3-digit groupings. In most cases, catch-all ‘other’ or ‘not elsewhere classified’ (nec) codes are included. In 1990, the first classification of occupations in the UK was produced. This was known as the Standard Occupational Classification (see ED Group of OPCS, 1990). SOC 90 was used for a variety of employment-related policy purposes, such as work-based training. Following a revised index to SOC 90 produced in 1995, a revised and updated SOC 2000 classification was produced (ONS, 2000). SOC 90 and SOC 2000 both classify occupations by skill level and skill content. The latter has nine major (1-digit), 25 sub-major (2-digit), 81 minor (3-digit) and 353 unit (4-digit) groups (ONS, 2000, pp. 21–36). The relevant SIC and SOC codes allow useful analyses to be undertaken of key trends across many of the sectors sponsored by the DCMS and its bodies. The employment focus of the SOC system enables analyses of employment at a detailed level. Thus Creigh-Tyte and Thomas (2001) and Creigh-Tyte, Creigh-Tyte and Thomas (2003) provide a comprehensive review of cultural employment trends in the late 1990s, building on the earlier analysis of Casey, Dunlop and Selwood (1996). The creative industries pose particular problems because, unlike the often well- established public-sector-based ‘cultural’ areas, they are mostly private service- sector-based activities which include cross-cutting specialist economic areas. As 164 A. Creigh-Tyte noted earlier, the CITF produced two successive Mapping Documents during Labour’s first term in government (see DCMS & CITF, 1998; DCMS & Ministerial Creative Industries Strategy Group (MCISG), 2001). However, even by the second Mapping Document the MCISG, which had succeeded the CITF, noted that

Although there has been some improvement in the provision of official statistics, the complex nature of the creative industries makes ‘scientific’ definition difficult, so there are still difficulties in producing robust data which separately identify the creative industries. (DCMS & MCISG, 2001, p. 0005) The second Mapping Document has not been updated, but the DCMS has produced a series of annual Creative Industries Economic Estimates (CIEE) based upon the ‘best- fitting’ SIC and SOC codes for the CITF’s 13 designated creative industries. The industrial and occupational classifications are summarized in Tables 4 and 5 respectively. The limitations of the SIC system in classifying the type of cross-cutting specialist activities which are particularly important in the creative industries are immediately apparent. Such activities include ‘Art/Antiques Trade’, ‘Crafts’, ‘Design’ and ‘Designer Fashion’. In the first two cases, neither the DCMS nor the Department of Trade and Industry (DTI) has proposed or employed any SIC-based definitions—see Table 4. At the occupational level, three 4-digit unit SOC codes are suggested for a combined ‘Design’ and ‘Designer Fashion’ category and the ‘Art/Antiques Trade’ category has no suitable SOC descriptive categories (see Table 5). For ‘Designer Fashion’, a Mintel definition is adopted encompassing: couture (the original designer market dominated by French-based international brands); international designers (a label usually dominated by one name); diffusion (designers producing ‘high-street’ ranges for specific stores); and high fashion (up-and-coming new designers, usually endorsed by celebrities) (DCMS & MCISG, 2001, p. 601). However, it is extremely difficult to delineate these activities within the official SIC and SOC data.

‘Scaling Factors’ and Creativity The official industrial classification issues are clearly the most significant in relation to all the main production dimensions, except perhaps employment, where the SOC system provides useful insights. The origins of the CIEE approach goes back to the deliberations of the CITF, which suggested that

in lieu of any dedicated SIC codes, the only way to measure ‘Designer Fashion’ is to estimate that it accounts for a tiny fraction (0.5%) of nine different clothing production codes. A similar approach based on estimated weightings of non-cultural 4 digit codes was taken for Architecture and the Art and Antiques market. (DCMS & ERCC, 2002a, p. 17) These ‘weightings’ or ‘scaling factors’ are clearly intended to reﬂect the proportions of the relevant SIC codes accounted for by ‘cultural activities’. As shown in Table 4, such ‘scaling factors’ are applied to seven out of the 11 (combined) creative industry groupings. However, in ‘Architecture’, ‘Art/Antiques Trade’ and ‘Designer Fashion’, Cultural Trends 165 Table 4 Creative Industries Economic Estimates: Assumption on Correspondence between Creative Industries and SIC Codes

CITF Mapping Document chapter Sector Standard occupation classiﬁcation (SIC)

1 Advertising 74.4 Advertising 2 Architecture 74.2 Architecturalab 3Art/Antiques Trade 52.48/9 Retail sale in specialized stores neca 52.5 Retail sale of secondhand goods in storesa 4 Crafts Majority of businesses too small to be picked in business surveys 5 Design No codes match this ﬁeld 6 ‘Designer Fashion’ Nine sub-sectors of clothing manufacturea 74.84 Other business activities neca 7 Video, Film, Music and 22.32 Reproduction of video recordinga Photography 92.11 Motion picture and video production 92.12 Motion picture and video distribution 92.13 Motion picture projection 74.81 Photographic activitiesa 9 and 10 Music and the Visual and 22.14 Publishing of sound recordings Performing Arts 22.31 Reproduction of sound recordinga 92.31 Artistic and literary creation and interpretation 92.32 Operation of arts facilities 92.34 Other entertainment activities neca 92.72 Other recreational activities neca 11 Publishing 22.11 Publishing of books 22.12 Publishing of newspapers 22.13 Publishing of journals and periodicals 22.15 Other publishinga 92.4 News agency activities 8 and 12 Software (including leisure) 22.33 Reproduction of computer mediaa and Computer Services 72.2 Software consultancy and supply 13 Radio and TV 92.2 Radio and television activities

Notes: aDenotes that only a proportion of this industry group is included to estimate the creative element. Details are available on request. bNumber of architecture business has been calculated using code 74.20/1 from the IDBR. Source: DCMS (2004b, Annex A)

‘scaling’ is applied to all of the constituent SIC categories. ‘Design’ and ‘Crafts’ have no appropriate SIC codes at all. ‘Scaling’ is applied reportedly on the basis of undocu- mented consultations with industry. The SIC sub-classes to which the ‘scaling factors’ are applied in the case of ‘Designer Fashion’ are set out in Table 6. While elements of ‘Designer Fashion’ may be present in most of the clothing production codes, it is far from clear that the application of a 166 A. Creigh-Tyte Table 5 Creative Industries Economic Estimates: Assumptions on ‘Best Fitting’ SOC 2000 codes for Creative Occupations

CITF Mapping Standard Occupation Classiﬁcation Document chapter Sector (SOC)

1 Advertising 1134 Advertising and public relations managers 3433 PR officer 3543 Marketing associate professional 2 Architecture 3121 Architectural technologists and town planning technicians 3Art/Antiques Trade None 4 Crafts 5491 Glass and ceramics makers, decorators and finishers 5492 Furniture makers, other craft woodworkers 5493 Pattern makers (moulds) 5494 Musical instrument makers, tuners 5495 Goldsmiths, silversmiths, precious stone workers 5496 Floral arrangers, florists 5499 Hand craft occs nec 8112 Glass and ceramics process operatives 9121 Mates to woodworking craftsmen/ women 5 and 6 Design and ‘Designer 2126 Design and development engineers Fashion’ 3411 Artists 3421 Graphic designers 7 Film and Video 3434 Photographers and audio-visual equipment operators 9 and 10 Music and the Visual and 3412 Authors, writers, journalists Performing Arts 3413 Actors, entertainers 3414 Dancers and choreographers 3415 Musicians 3416 Arts officers, producers and directors 11 Publishing 3431 Journalists, newspaper and periodical editors 5421 Originators, compositors and print preparers 5422 Printers 5423 Bookbinders and print finishers 5424 Screen printers 8 and 12 Software (including leisure) 1136 IT/Communications managers and Computer Services 2131 IT professionals 13 Radio and TV 3432 Broadcasting associate professionals 5244 TV, video and audio engineers

Source: DCMS (2004b, Annex A) Cultural Trends 167 Table 6 ‘Designer Fashion’ on the Creative Industries Task Force Deﬁnition SIC (92), 1997 revision

Section D Manufacturing

Sub-sections DB Manufacture of textiles and textile products DC Manufacture of leather and leather products K Real estate, renting and business activities Divisions 17 Manufacture of textiles 18 Manufacture of wearing apparel; dressing and dyeing of fur 19 Tanning and pressing of leather; manufacture of luggage, handbags, saddlery, harness and footwear 74 Other business activities Classes 17.7 Manufacture of knitted and crocheted articles 18.1 Manufacture of leather clothes 18.2 Manufacture of other wearing apparel and accessories 18.3 Dressing and dyeing of fur: manufacture of articles of fur 19.1 Tanning and dressing of leather: manufacture of luggage, handbags, saddlery, harness and footwear 19.2 Manufacture of luggage, handbags and the like, saddlery and harness 19.3 Manufacture of footwear 74.8 Miscellaneous business activities not elsewhere classiﬁed (nec) Sub-classes ‘Designer Fashion’ is estimated by applying a 0.5 of one percentage point ratio to the total estimates for: 17.71 Manufacture of knitted or crocheted articles 17.72 Manufacture of knitted or crocheted pullovers, cardigans, jerseys, waistcoats and similar articles 18.10 Manufacture of leather clothes 18.21 Manufacture of work wear 18.22 Manufacture of other outerwear 5-digit 18.22/1 Other men’s outerwear 18.22/2 Other women’s outerwear 18.23 Manufacture of underwear 5-digit 18.23/1 Men’s underwear 18.23/2 Women’s underwear 18.24 Manufacture of other wearing apparel and accessories nec 18.24/1 Hats 18.24/3 Cut, make and trim (CMT) for clothing manufacturers 18.24/9 Other wearing apparel and accessories nec 18.30 Dressing and dyeing of fur; manufacture of articles of fur 19.10 Tanning and dressing of leather 19.20 Manufacture of luggage, handbags and the like, saddlery and harness 19.30 Manufacture of footwear Plus 0.25 (25%) share in 74.84 Other business activities (nec)

Sources: ONS (1997), DCMS and ERCC (2002b) single uniform ‘scaling factor’ captures this with any degree of accuracy. At the most basic level, much more attention and commercial focus is placed by the designer sector on womenswear products. Similarly, workwear manufacture may be less likely to include any substantial ‘Designer Fashion’ element. Equally, ‘Designer Fashion’ 168 A. Creigh-Tyte elements may not reflect the overall core industry average. For example, ‘Designer Fashion’ enterprises could quite conceivably be smaller than the average clothing business, so their share of value added or employment would be less than their share in total business numbers. A single crude ‘scaling factor’ cannot reflect these differences. The assumption that designer enterprises are the same as clothing enterprises more generally is distorting. DCMS itself acknowledges that this ‘is not necessarily a good match’ and indeed DCMS do not usually refer to the 5-digit sub-classes available within the SIC, simply because actual data are not available for these sub-classes. Thus distinctions between men’s and women’s other outerwear manufacture, 18.22/1 and 18.22/2 respectively, are not, in practice, possible using the published official statistics (DCMS & English Regional Cultural Consortia (ERCC), 2002b, p. 70). The weakness of the relationship between ‘Designer Fashion’ and the much broader 17.71–19.30 SIC manufacturing groups is highlighted by the fact that all the other ‘scaling factors’ employed by the CITF are 25 or 50 per cent, with the exception of the equally slippery ‘Art/Antiques Trade’ creative industry, where 5 per cent ratios, themselves 10 times the level used for ‘Designer Fashion’, are applied to both of the constituent sub-classes, in this case 52.48/9 and 52.50 in the retail trade division. In other words, the application of the ‘scaling factor’ would also appear arbitrary and inconsistent. Even the DCMS itself is not completely consistent on the ‘scaling factors’ used. Thus, at one point, ‘Designer Fashion’ is defined at the 4-digit SIC level as a ‘... tiny fraction (0.5%) of nine different clothing industry production codes’ (DCMS & ERCC, 2002a, p. 17). However, the other wider definition usually employed (as in Table 6) includes ‘17.71, 17.72, 18.0, 18.10, 18.21, 18.22, 18.23, 18.24, 18.30, 19.30, these codes cover manufacturing of clothes and footwear, within which are “Designer Fashion” activities 74.84 other business activities nec (DCMS & ERCC, 2002b, p. 69)’. Moreover, in the application of ‘scaling factors’ applied to other business activities, nec74.84, the ‘factor’ is itself unclear. Most sources suggest a 25 per cent weight: ‘... creative industry share of total class (%) ...17.71–19.30 0.5 ...74.84 25’ (DCMS & ERCC, 2002b, p. 70). Although later a much lower weight is indicated: ‘... creative industry share of total class (%) 17.71–19.30 0.5 ... 74.84 (no longer used) 2.0’ (DCMS, 2004c, p. 67). Minor variations in the ‘scaling factors’ used in estimates of the economic value of ‘Designer Fashion’ have a potentially large absolute effect because the core sectors to which the weights are applied are so large. The ONS Annual Business Inquiry has since 1998 provided integrated employment and accounting information on 13 main variables for business across the UK economy. (See Jones (2000) for a discussion of the Annual Business Inquiry’s development.) Data for selected economic variables in 2002, which is the latest year available at the time of writing (December 2004), for the clothing production and other business activities codes are summarized in Table 7. The UK’s clothing production sector has shrunk steadily in recent years. For example, the gross value added in ‘Manufacture Cultural Trends 169 Table 7 Selected ABI estimates for Clothing Production and Other Business Services nec, 2002

Total Gross value employment Number of Total turnover added average during enterprises (£ million) (£ million) year (000s)

Clothing production 17.71–17.72 505 697 306 17 18 5,383 4,657 1,612 73 19 898 1,541 571 17 Total clothing production 6,786 6,895 2,489 107 Other business activities nec 74.84 77,431 24,979 11,642 350

Source: ONS ABI Spreadsheets kwww.statistics.gov.uk/variable_info.aspl, June 2004

of Wearing Apparel; Dressing and Dyeing of Fur’ (SIC Division 18) fell from over £2.7 billion in 1998 to £1.6 billion in 2002. In contrast, gross value added in the Other Business Activities nec category (74.84) rose from £10 billion to £11.6 billion over the same 1998–2002 period. Nevertheless, even Clothing Production, with almost 7,000 enterprises turning over almost £7 billion and with gross value added approaching £2.5 billion and average employment of 107,000, constitutes a substantial economic sector, which is linked to the ‘Designer Fashion’ creative industry in obscure and complex ways which cannot be adequately captured by a single tiny ‘scaling factor’ share. The CITF/MCISG have not updated the 2001 Mapping Document. However, the DCMS has produced three successive annual CIEE reviews. The latest of these appeared in the summer of 2004 and provides the DCMS’s best estimates of the development of the creative industry sector up to 2003 (DCMS, 2004b). DCMS’s CIEE provide various time-series data based on the ‘scaling factor’ approach for ‘Designer Fashion’ for various time periods back to 1995. In several cases, ‘Design’ estimates based largely on the design consultancies sector are merged with those for ‘Designer Fashion’ (see Table 9). However, the rationale for this, other than the shared difficulties in using standard statistical definitions, is unclear. (See British Design Initiative & Design Council (2002, 2003) for a valuation of the ‘Design’ sector.) CIEE data for gross value added, export revenues and business numbers are summarized in Table 8, while employment trends for ‘Design’ and ‘Designer Fashion’ and the creative industry as a whole since 1995 are provided in Table 9. It is apparent that ‘Design’ or, in this particular case, the design consultancies, constitutes a far larger, if less clearly defined, sector. For example, in 2001 ‘Design’ value added and exports were 20 times and two and a half times respectively greater than corresponding values for ‘Designer Fashion’. The rationale for, and utility of, a single combined employment total as presented in Table 9 is therefore unclear. 170 A. Creigh-Tyte Table 8 Economic Trends in ‘Design’ and ‘Designer Fashion’

1996 1997 1998 1999 2000 2001 2002 2003

Gross value ‘Designer Fashion’ — 280 270 300 360 320 300 — added (£ million) Designa — — — — 6,500 6,760 5,900 — Exports (£ million) ‘Designer Fashion’ — — 350 — — 390 — — Design — — — — 1,000 1,000 1,400 — Number of ‘Designer Fashion’ 1,400 1,400 1,300 1,300 1,300 1,300 1,300 1,300 businesses

Note: aDesign related to turnover of design consultants Source: DCMS (2004b, Tables 1a, 2 and 4, pp. 5–7)

Some of the CIEE economic data mesh quite well with the material in the original CITF Mapping Documents, especially if we allow for the fact that the CITF data are often only vaguely dated, and as the CITF itself stressed, the two Mapping Documents ‘...are not directly comparable’ (DCMS & MCISG, 2001, Appendix, p. 002). The CIEE export estimates appear to have a CITF base, e.g. £350 million for ‘Designer Fashion’ in 1998 and £1,000 million for ‘Design’ in 2000 and 2001 (see Table 10). However, the CITF’s combined estimate for employment is only 35,000 in the 1998 Mapping Document, whereas the lowest CIEE estimate of employment in ‘Design’ and ‘Designer Fashion’ is 79,500 (in 1995). The CIEE employment estimate apparently sums the three ‘Design’ 4-digit SOC codes (see Table 5). For 2002, the CIEE estimates value added in ‘Designer Fashion’ as £300 million from 1,300 businesses. This represents 2.1 and 1.5 per cent respectively of the gross value added and number of enterprises in clothing production and other business activities nec combined according to the ONS’s Annual Business Inquiry (ABI) data for 2002 (see Table 7). However, these shares do not correspond directly to any

Table 9 Creative Employment Trends in Great Britain

‘Design’ and ‘Designer Fashion’ Total creative employees

Total employment (000s) (summer quarter) 1995 79.5 1499.4 1996 85.5 1503.7 1997 80.7 1572.1 1998 88.8 1669.3 1999 93.5 1735.1 2000 98.5 1773.5 2001 103.0 1828.1 2002 115.0 1830.7 2003 113.2 1878.8 Annual growth 1995–2003 5% 3%

Source: DCMS (2004b, Table 3, p. 7) Cultural Trends 171 Table 10 Mapping the Creative Industries: The 1998 and 2001 Documents

1998 2001

Revenue (£ billion) Design 12.0 26.7 ‘Designer Fashion’ 0.6 0.6 Employment (000s) Design 23 76 ‘Designer Fashion’ 12 12 Exports (£ million) Design 350 1,000 ‘Designer Fashion’ 350 350

Source: DCMS & MCISG (2001, Appendix, pp. 003–005)

combination of the admittedly ambiguously deﬁned ‘scaling factors’ suggested for quantifying ‘Designer Fashion’ by the CITF (see Table 6).1 The potentially saving grace of the CIEE approach is the continuity it provides across the three annual reports produced to date. This allows trends over time in particular sectors to be assessed, at least compared to creative industries as a whole, even if the exact accuracy of the data for individual years is questionable. Thus, as shown in Table 9, it seems that employment in ‘Design’ and ‘Designer Fashion’ is growing mark- edly faster on average than for the creative industries as a whole.

Alternative Approaches to Documenting the Creative Industries The ‘Creative Sector’ Production Chain The ONS’s UK Input–Output Analyses Annual Supply and Use tables identify supply from one industry and use within the economy. Much of this supply forms ‘intermediate goods’ used in the next stage in the production cycle. These ‘intermediate goods’ do not form part of the gross domestic product (GDP). Mahajan (2004) provides an overview of the ‘Creative Sector’ on an input–output basis from 1992 to 2002. This approach uses SIC (92) industry data allocated to the creative industry sector by the CITF Mapping Document of 2001. However, in a freely acknowledge departure from CIEE practice, no attempt is made to sub-divide the relevant SIC categories into ‘creative’ and other elements (Mahajan, 2004, p. 43). Thus all of the nine clothing production 4-digit headings and other business activities nec (74.84) documented within the ABI are allocated to the ‘Clothing’ heading within the input–output analyses. Not surprisingly, this produces a very different series of economic estimates. ‘Cloth- ing’ formed the largest proportion of creative products’ supply at almost £81.9 billion (35 per cent) out of a total of over £234.4 billion in 2002. Likewise, the ‘Clothing’ input–output industry category’s share of gross value added in 2002 reached £13.2 172 A. Creigh-Tyte billion (16 per cent) out of a total of almost £80.9 billion. In turn, the input–output ‘Creative Sector’ represented 8.7 per cent of total UK gross value added of £926.3 billion in 2002.2

The ‘Culture Cycle’ During 2000/01 the English Regional Cultural Consortia (RCC) encountered a range of difficulties in trying to assemble a quantitative picture of ‘culture’ in their regions as an input to regional development agendas.3 This led the RCC and the DCMS to sponsor a project which would give a ‘blueprint’ of definitions and concepts for the cultural sectors to provide a reference source on key data sources and practical guidance on best practice for quantitative research in the cultural sectors. The resulting Regional Cultural Data Framework user’s guide and its accompanying technical report appeared in the autumn of 2002 (DCMS & ERCC, 2002a, 2002b). The Regional Culture Data Framework provides useful guidance on research data sources and a worked example of data extraction, but our focus here is on the Regional Culture Data Framework’s analytical framework proposals. The Regional Culture Data Frame- work adopts a pragmatic if simplistic definition of the cultural sector as simply ‘... those activities for which DCMS has responsibility’ (DCMS & ERCC, 2002a, p. 9). The Regional Culture Data Framework reflects the original rings in the mandala sketched out in Labour Party (1997), and the cultural industries production system operationalized by Pratt (1997), in advocating a definition of the cultural sector representing the sum of activities and resources involved in a multiple phase, ‘Cycle’ of creation, making, dissemination, exhibition/reception, archiving/presentation and education/understanding for cultural products and cultural services. (See Pratt (2004) for a discussion of the cultural industries production network model.) This comprehensive and ambitious approach goes well beyond the creative industries to include a wide range of publicly subsidized sectors as well as sport and tourism. Thus the DCMS’s sectors were grouped into seven ‘Cultural Domains’ of Visual Art, Performance, Audio-visual, Books and Press, Sport, Heritage and Tourism. Within the Regional Culture Data Framework, and subsequently the re-titled DCMS Evidence Toolkit framework, ‘Design-based activities’ such as fashion, interior, building (aka architecture), graphic and product design all fall within the ‘Visual Art’ area, which is one of seven ‘cultural domains’ (DCMS & ERCC, 2002a, p. 11). However, the actual activities undertaken cross many of the 13 creative industries and often spread across the various ‘domains’ (the Arts sector raises similar issues). Moreover, as the DCMS and ERCC acknowledge, there are strong cross-domain links between ‘Visual Art’ and ‘Heritage’, as artefacts created in the former domain are exhibited, archived, conserved and restored within the museums and archives of the latter domain (DCMS & ERCC, 2002a, p. 11). In presenting a definitional matrix for the six functions across the seven domains, the Regional Culture Data Framework made little attempt to develop the SIC code mapping for the design-based activities. While the ‘Creation’ function includes 1992 SIC code 74.20/1 ‘Architectural activities’, the entry for ‘Other design activities Cultural Trends 173 (fashion, graphic, interior, product)’ simply notes NFW—‘Needs Further Work’ (DCMS & ERCC, 2002a, p. 18; 2002b, p. 22). In any case, readers of the Regional Culture Data Framework will search in vain for actual data—even at national level. Like many ‘blueprints’, the structure remains unbuilt and the data cells unpopulated. In this case researchers on some of the con- ventional SIC code-based creative industries are in a relatively privileged position given the annual CIEE reviews. The 1992 SIC was revised in 1997 to introduce 111 new, most disaggregated 5-digit sub-classes and the balancing ‘not elsewhere classified’ and ‘other’sub-classes were renumbered. The ONS’s introduction of further limited revisions in the SIC 2003 codes and con- tinued consultation and development work by the DCMS since the Regional Culture Data Framework’s appearance in late 2002, culminated in the publication of a re-titled DCMS Evidence Toolkit (DET) in the summer of 2004 (DCMS, 2004c). The DCMS Evidence Toolkit provides revised definitions of all the sub-sectors within the DCMS’s policy remit, including the creative industries. Like the Regional Culture Data Framework, it incorporates: . the inclusion of the whole ‘Culture Cycle’ production chain or Cycle (including Creation, Making, Dissemination, Exhibition/Reception, Archiving/Preservation and Education/Understanding); but it also features: . the adoption of the SIC 2003 which includes some new 5-digit, most disaggregated detailed codes; and . the abandonment of the ‘scaling factors’ applied since the CITF’s original ‘mapping’ work. The ‘Designer Fashion’ and wider ‘Design’ creative industries are perhaps the most striking examples of the potential implications of the new DCMS Evidence Toolkit approach. For ‘Designer Fashion’ the DCMS Evidence Toolkit suggests the use of the new 74.87/2 sub-class ‘Speciality designer activities’ in the ‘1. Creation’ function of the ‘Visual Arts’ domain (see Table 11). Within the ‘Other business activities nec’ (74.87) 4-digit class, the old sub-class 74.84/1 covering ‘credit reporting and collection agency activities’ (ONS, 1997,

Table 11 Design Activities in SIC (2003)

74.87/2 Specialist design activities This sub-class includes: † fashion design related to textiles, wearing apparel, shoes, jewellery, furniture and other interior decoration and other fashion goods as well as other personal or household goods † activities of interior decoration designers † activities of graphic designers This sub-class excludes: † machinery and individual plant design cf. 74.20/5 † display of advertisements and other advertising design cf. 74.40/2

Source: ONS (2002, p.158) 174 A. Creigh-Tyte p. 244) was renumbered as 74.87/1 under SIC 2003, but still grouped with ‘speciality design activities’ (74.87/2), ‘activities of exhibition and fair organisers’ (74.87/3), ‘activities of conference organisers’ (74.87/4) and ‘other business’ activities nec (74.87/9) (ONS, 2002, p. 158). Sub-class 74.87/3 ‘activities of exhibition and fair organisers’ appears in the ‘4. Exhibition/Reception’ function and presumably includes (among many other things) fashion exhibitions and parades. However, there are no specific classifications for any of the other ‘Functions’ within the ‘Culture Cycle’—‘2. Making’, ‘3. Dissemi- nation’, ‘5. Archiving/Preservation’ and ‘6. Education/Understanding’ (see DCMS, 2004c, Table 2, p. 28). Even for the broad ‘Visual Arts’ domain there are in practice no appropriate classifications at all for Domains 2, 5 and 6. The framework simply notes ‘Needs Further Work’ in these cells. However, it is distinctly disappointing that ‘Design’ (let alone ‘Designer Fashion’) is not even mentioned under the ‘Education/Understanding’ area of Visual Arts, in spite of the fact that these disciplines were among the most rapidly growing subjects to be studied at least in the higher education level4 (DCMS, 2004c, Table 2, p. 28). At least ‘Education/Understanding’ warrants a ‘Needs Further Work’ entry, whereas the ‘Archiving/Preservation’ function of ‘Visual Arts’ apparently warrants no comment at all. The ‘Culture Cycle’ approach advocated in the Regional Culture Data Framework and DCMS Evidence Toolkit has major implications for the scale of the DCMS’s sectors, especially under the more developed DCMS Evidence Toolkit proposals for the supply chain. The most dramatic examples lie in the sectors for which more adequate SIC codes exist, rather than cases such as ‘Crafts’, ‘Design’ and ‘Designer Fashion’ where even the core ‘Creation’ function is difficult to quantify using official statistics. Thus, for example, within the Audio Visual ‘domain’ code 92.2 ‘radio and television activities’ clearly forms part of the ‘Creation’ core. However, within the DCMS Evidence Toolkit’s ‘Making’ function code 32.1 ‘Manufacture of electronic valves, tubes and other electronic components’ together with 32.20/2 ‘Manufacture of television and radio transmitters, etc.’ and 32.3 ‘Manufacture of TV and radio valves, sound or video recording or reproducing apparatus and associated goods’ all appear. Thus electronic components, the making of TV and radio transmitters and manufacture of consumer entertainment equipment all fall within the ‘Culture Cycle’. Manufacture of mass-production TVs and radios apparently has a ‘Culture’ dimension. Moreover, the logic of the ‘Culture Cycle’ is extended to include under ‘Dissemina- tion’ the distribution of such main market consumer goods. Hence both 51.43/9 ‘Wholesale of radios and televisions; wholesale of electrical appliances not elsewhere classified’ and 52.45 ‘Retail sale of electrical household appliances and radio and television goods’ are included, so that Comet and Dixons, for example, take their place within the ‘Culture Cycle’ (DCMS, 2004c, p. 29). Attempts to boost the size of the creative industries and the cultural sector more generally have long been an often-deplored feature of ‘research’ in these fields. Size is used as a tool for advocating purposes. However, most recent research has sought Cultural Trends 175 to provide evidence on ‘what works’ for advocacy policy development, rather than relying on the simplistic and fallacious argument that (large) size as such justifies government support. If the DCMS Evidence Toolkit approach represents a retreat from this approach it constitutes a distinct backward step.5 The DCMS Evidence Toolkit expresses the hope that eventually all the activities within an SIC code could be a specific part of the cultural sector, but realistically notes ‘... There is much work to be done before this becomes reality’ (DCMS, 2004c, p. 14). Over the medium term there is some prospect of improvements in the documen- tation of the creative industry sectors. A new SIC revision is scheduled for 2007, and, while 4-digit classifications must mesh with the EU’s Industrial Classification (NACE) over which individual Member States clearly have only limited influence, the most detailed 5-digit-level sub-class classification remains at each country’s own discretion. However, even if greater detail on the poorly documented creative industries becomes possible under SIC 2007, actual data will not become available until 2008 or 2009. As noted earlier, even at the end of 2004 there are no data available for the new ‘Designer Fashion’ 5-digit SIC 2003 sub-class recommended in the DCMS Evidence Toolkit. 6 The ABI data for 2002 were only released in June 2004 and are based on SIC 92/97. Thus it remains to be seen how the data on ‘Designer Fashion’ on the new DCMS Evidence Toolkit basis compare to the established ‘scaling factor’ approach used in the CIEE report series to date.

Specialist Studies of ‘Designer Fashion’ in the UK The 1990s Given the difficulties in producing reliable analyses of creative industries such as ‘Design’ and ‘Designer Fashion’ which cut across a variety of official statistics categories outlined above, an obvious alternative approach is to undertake specialized studies of the sectors concerned. The DTI shares ‘sponsorship’ of the creative industries with the DCMS and is primarily responsible for sponsoring ‘Designer Fashion’ and ‘Design’. Thus the 1998 Mapping Document provides basic estimates of ‘Designer Fashion’s’ economic scale, noting that ‘all figures derive from ad hoc DTI research based on industry interviews’ (DCMS & CITF, 1998, p. 044). However, the bibliography of the Mapping Document lists Coates (1997) as the sole identifiable evidence source for ‘Designer Fashion’ (DCMS & CITF, 1998, p. 109). By the time of the 2001 Mapping Document some limited further analysis on ‘Designer Fashion’ had been produced for the DTI (Cheshire, 1998), but given that this report appeared at the same time as the first Mapping Document, it is hardly surprising that most of the detailed estimates relate to 1996 or 1997. Both Coates (1997) and Cheshire (1998) draw historical data from an earlier Kurt Salmon Associates (KSA) report commissioned by the British Fashion Council (KSA, 1991). KSA’s methodology is unclear and Coates (1997) was primarily intended as a 176 A. Creigh-Tyte ‘self-help’ manual for fashion designers rather than a research report. Since Cheshire (1998) draws heavily on both KSA (1991) and Coates (1997), it is reasonable to conclude that none of the detailed customized data on ‘Designer Fashion’ in either of the CITF’s Mapping Documents has a clear robust source.

‘Designer Fashion’ in 2001 In April 2002 the Malcolm Newbery Consulting Company (hereafter ‘Newbery’) was commissioned by DTI and the British Fashion Council to undertake a further specialized research project. Like Coates (1997), Cheshire (1998) and KSA (1991) before him, Newbery’s study had a complex remit. In fact, Malcolm Newbery was the main author of the KSA review of the industry in 1990. The new Newbery study aimed to carry out: ‘An investigation producing quantified outputs, of the impact of the “Designer Fashion” industry on the wider UK textiles and clothing industry’ (Newbery, 2003, p. 4). However, the primary focus was on the business support needs of ‘Designer Fashion’, the sources of the business support and mapping current gaps in support provision.7 Of the £75 million UK designer income in 1990, some £60 million came from direct sales (wholesale and retail), with the remaining £15 million from licence income, at around one-eighth of the licensed sales of £125 million returning as income to UK designer licensees. By 2001, direct sales were estimated at £625 million, with licence fees accounting for about £75 million, 11 per cent, of total income. Taken literally, the data in Table 12 show compound real sales growth of over 22 per cent per annum for 1990 to 2001, since there was little, if any, inflation in clothing prices over the period. What Newbery terms the ‘Designer Universe’ of ‘Design business selling ... up market range’,8 numbered 791 businesses with sales of £611 million. Of these, the smallest 500 were often craft shops with median annual sales of £35,000 and total business sales of £18 million (Newbery, 2003, Appendix 4, p. XXI). While Newbery’s report represents a significant improvement in transparency in methodology compared to earlier specialist studies, with over 50 scoping expert interviews and published questionnaires, Newbery’s results are based on responses from a sample of ‘over 200’ designers, 25 retailers and 25 UK manufacturers. However, only 70 questionnaire respondents are listed (Newbery, 2003, Annexe B).

Table 12 Independent Fashion Designer Sector in the UK

1990 1996 2001

Total UK designer income (£ billion) 75 500a 700

Note: aCheshire (1998) valued this at £600 million (see Table 10), but Newbery (2003) argues that this ﬁgure does allow for the variety of sales routes and adjusts the estimate downwards to £500 million. Source: Newbery (2003, Appendix 2, pp. IX–XI) Cultural Trends 177 So, in spite of distinct improvements in transparency, the exact derivation of the ‘Designer Universe’ ﬁgures and their reliability remains unclear. This is especially important because Newbery produces estimates that are considerably larger than some of the CIEE ‘scaling factor’ equivalents. Thus designers provide more employment pro rata than any other part of the clothing industry, accounting for 6 per cent of the UK clothing workforce (Newbery, 2003, p. XIV).9

Anglo–Italian Comparisons Given time and resources, insightful analyses of the relevant spheres are possible. A recent DTI-ﬁnanced comparative study focused on the reasons underlying Italy’s relative success in clothing and textiles.10 Owen and Cannon Jones (2003) report an important two-year benchmarking study of woollen and worsted fabric and men’s tailored suits which documents the Italian industry’s successful focus on geographic clustering, quality enhancement and value-added strategies.

The Regional Dimension? In a project to produce data on the cultural industries funded by the DCMS, Rood- house (2003) sought to provide empirical data on ‘Designer Fashion’ in the North West and Yorkshire and Humberside regions of England. The regional focus, especially on areas outside London and the South, which is normally viewed as the ‘home’ of the UK’s ‘Designer Fashion’ sector, is unusual. However, while pointing out the limitations of the SIC/SOC codes, Roodhouse veers to the other extreme in adopting an alternative approach to provide ‘a realistic and “democratic spectrum”’ (Roodhouse, 2003, p. 3). Essentially, this is a self-defined approach with survey respondents in clothing production and its related and periph- eral activities,11 e.g. fashion photographers defining their own ‘designer’ status and the segments of the market in which they operate. This inclusive approach embraces manufacturers with a tenuous link to Newbery’s ‘Designer Universe’ (Roodhouse, 2003, pp. 3–4). However, even for Roodhouse such ‘democratic’ self-inclusion also has its limits. While education as a source of employment and manufacturing are included in ‘Designer Fashion’, ‘... perfume design, modelling, magazine production, and high street retailing are excluded, as the active “Designer Fashion” interaction is implicit’. Apparently this is because ‘the guiding principle has still been to capture as much design activity as possible, hence ignoring high street retail outlets’ (Roodhouse, 2003, p. 4).12 Some 1,000 businesses across the two regions were sampled using a postal questionnaire, but this produced only a 20 per cent response rate; so, 204 businesses were in the achieved sample (Roodhouse, 2003, p. 41). Inspection of the respondent sample underlines that respondents are predomi- nantly small clothing businesses serving regional or UK markets. Micro-businesses with under 10 people employed dominate the respondent data base, with 43 per cent having five or less full-time employees, and with 41 per cent of main 178 A. Creigh-Tyte customers in the regions concerned and a further 50 per cent located elsewhere in the UK. Roodhouse (2003) produces results for a range of variables for the financial year ending in 2001, but few overall estimates of the sector’s size and significance. Given the response rate, that is probably a reasonable approach, but does little to illuminate the interactions between the ‘Designer Fashion’ creative industry and the wider clothing production sector or ‘Designer Fashion’ and its related sectors.

Conclusions In the eight years since the election of the Labour government in the UK, there has been a significant interest in stated policy towards the cultural sector, and the creative industries in particular. At the same time, the UK government has strived to place ‘what works’ rather than ideology at the heart of policy development and implementation.13 Policy towards many of the creative industries is handicapped by a limited evidence base, but there are particular issues surrounding sectors such as ‘Designer Fashion’ which do not have dedicated SIC codes. In these cases basic economic data on the sector have to be developed. The use of ‘scaling factors’ applied to much wider SIC codes, as originally suggested by the CITF, may have served as a stop-gap solution and did allow the first overall estimates of activity in the creative industries to be published in 1998. However, this approach is no longer adequate in ‘Designer Fashion’. The rationale for the ‘scaling factors’ applied to clothing production and other business activities nec codes is unclear, and indeed the weights themselves differ according to the source used. Over recent years the ONS and DCMS have explored alternative models for estimat- ing economic variables in both the creative industries generally and ‘Designer Fashion’. The ONS’s input–output approach clearly includes large sections of clothing production which have a low design content and consequently tends to overestimate the sector’s size. The ‘Culture Cycle’ production chain approach advocated by the DCMS in both the Regional Culture Data Framework and DCMS Evidence Toolkit ‘blueprints’ embraces six functions stemming from ‘Creation’ in broad conformance with the ‘rings’ surrounding the creative core of the Labour Party’s original mandala. However, in the creative industries where relevant SIC codes are available, literal adherence to the approach would result in inflated and largely meaningless estimates of economic scale. In the case of the audio-visual ‘domain’, this would include UK- based radio and television manufacturing industry and retailers like Dixons. This is the type of inflation found in the past in numerous ‘advocacy’ documents on the economic impact of culture, and of which independent commentators are increasingly sceptical. Cultural Trends 179 For poorly documented sectors like ‘Designer Fashion’ such ‘inflation’ is, of course, simply not possible, even if it was desirable. Another feature of the Regional Culture Data Framework and DCMS Evidence Toolkit ‘blueprint’,the use of detailed 5-digit SIC codes, may eventually allow relatively meaningful official statistics to be developed. However, these changes will take time to implement. At the time of writing (December 2004), no data at the 5-digit sub-class level are available. The latest ABI 4-digit data for 2002 were only published in June 2004. The revision of the United Nations and EU statistical classifications in 2007 offers prospects for further improvements in coverage of the dynamic creative industry sectors, at least in the medium term, especially at the 4- and 5-digit levels where there is EU and UK discretion. However, actual data for 2007 are unlikely to appear until 2008 or 2009. Given the clear inadequacies of the official statistical data, there is a strong case for specialized focused studies in ‘Designer Fashion’. Unfortunately, progress since 1990 on this front has also been modest. Successive studies undertaken by KSA (1991), Coates (1997) and Cheshire (1998) all had a divided focus, obscure methodology, sampling selection problems and apparently limited response rates. A subsequent study in the North West and Yorkshire and Humberside regions showed similar obscurity in the definitions used, potentially covering almost all clothing production, and achieved a low (20 per cent) response rate. The Anglo–Italian comparative study by Owen and Cannon Jones (2003) provides useful benchmarking results on two narrowly defined clothing and textile sub-sectors. Newbery focused heavily on business support for ‘Designer Fashion’ in 2001, but did represent a distinct improvement in transparency by providing copies of the survey instruments used and produced 70 responses (Newbery, 2003). The ‘Designer Universe’ of 2001 as scoped by Newbery appears distinctly different from that derived from the DCMS’s ‘scaling factors’. Thus, for example, around 6 per cent of total employment in clothing production derives from the designer sector, hardly comparable to the 0.5 of a percentage point ‘scaling factor’ applied by the DCMS. When the DCMS mobilizes enough resources to undertake detailed research on sponsored sectors that do not fit simply into standard SIC categories, impressive results can be achieved. ‘Tourism’ provides an interesting comparison. Thus last year saw the publication of a detailed independent review of tourism statistics (Allnutt, 2004), the UK’s first Tourism Satellite Account (Jones et al., 2004) and a five-year tourism development strategy for England (DCMS, 2004a). However, eight years after the Labour Party launched its strategy document for ‘Designer Fashion’ and the other creative industries, a robust basis for assessing the scale and evolution of the British ‘Designer Fashion’ sector remains elusive. British fashion designers who lead the sector are considered amongst the most talented and creative in a global industry. They are frequently poached by leading French and Italian fashion houses, e.g. John Galliano to Dior. However, without a reliable data base, discussions about the sector’s future and the targeting of customized support14 will continue to take place in a vacuum. 180 A. Creigh-Tyte Acknowledgements The author wishes to thank the two anonymous referees and the editor for very helpful comments upon the first draft of this article. These suggestions have either been incor- porated in the final text, or will be taken forward in work in progress.

Notes [1] For 2002 ABI data, applying 0.5 of a percentage point ‘scaling factor’ to Clothing and 2.5 per cent to Other business activities nec yields a gross value added estimate of £303 million, almost identical to the CIEE’s £300 million—see Table 8. However, the same ‘scaling factors’ applied to enterprise numbers yield 1,970 businesses against the CIEE estimate of 1,300. Moreover, as noted earlier, a 2.5 per cent weighting for Code 74.84 does not appear in the various discussions of methodology produced by the DCMS. [2] Thus ONS’s 8.7 per cent share in gross value added in 2002 exceeds the 8 per cent estimate in the CIEE (see DCMS, 2004b, p. 6). [3] The tendency for creative industries to be subject to competitive and organizational pressures such that they frequently concentrate together in clusters or industrial districts is noted by Power & Scott (2004). Rantisi (2004) illustrates the relationships between cultural production and the city in a case study of the fashion industry in New York, and Vervaeke and Le Febvre (2002) illustrate the complexity of design and inter-firm relationships and shifting balances of power across the textile industry in the Nord-Pas de Calais region. [4] Between 1995/96 and 2002/03 the total number of students at undergraduate and post-graduate level grew by 24 per cent (to over 2,175,115). Creative arts and design student numbers grew by 52 per cent (to 132,625) and Design Studies by 32 per cent (to 53,615). In 1998, 9,600 first degrees were awarded in design, representing an increase of almost one-third since 1994/95. Some 91 fashion courses were operating in 2002, producing around 3,000 graduates per annum, according to Malcolm Newbery (Newbery, 2003, p. 38). In 2001–02 there were a total of over 20,000 first-year design students and 3,000 overseas first-year design students on British courses (Design Council, 2004). [5] Johnson and Thomas provide a valuable review of ‘impact’ studies across the arts sector and a discussion of the limitations of this approach to date. (See especially Johnson & Thomas, 2001, pp. 212–215.) [6] In the June 2004 data release, the most detailed ABI data publicly available on the ONS’s website are at the 4-digit level using the SIC 92/97 codes. [7] This focus reflected recommendations made by the Textile and Clothing Strategy Group (TCSG) (see TCSG, 2000). [8] Given its joint British Fashion Council sponsorship, it is not surprising to find that a fairly ‘elitist’ definition of the designer industry was adopted as: Individuals or teams that combine creativity and originality to produce a clothing collection with a specific or ‘signature’ identity, exemplified by, but not restricted to, the type of company that participates at international trade shows such as London Fashion Week and its equivalents. Fashion designers may produce diffusion lines in addition to their ‘flagship’ collections and range from established designers with an international reputation or ‘brand’ to ‘cutting edge’ newcomers. (Newbery, 2003, p. 5) [9] Newbery estimates overseas designer sales of £390 million in 2001, equivalent on some assumptions to 9 per cent of all UK clothing exports (see Newbery, 2004, pp. XIV–XV), and this £390 million export revenue figure is in turn used in the CIEE (see Table 8 above). The designer sector is growing; this is in sharp contrast to the UK clothing manufacturing sector as a whole (see Newbery, 2003, p. 43). Cultural Trends 181 [10] For example, Italy’s exports of men’s suits in 2000 were six times those achieved by the UK (Owen & Cannon Jones, 2003, p. 49). [11] Eight of the nine usual clothing production SIC classes appear to be included, but 18.10 ‘Manufacture of leather clothes’ is excluded (see Roodhouse, 2003, p. 3). [12] This delineation contrasts with the wider ‘Culture Cycle’ approach advocated in DCMS and ERCC (2002a, 2002b), which includes retail and wholesale distribution. [13] The need to considerably strengthen the DCMS’s research and science base is highlighted in the DCMS Science Review completed during 2004 (see Office of Science and Technology, 2004, pp. viii–xi). [14] Newbery (2003) emphasizes that a ‘one size fits all’ approach to supporting the UK ‘Fashion Designer’ sector is not applicable and support schemes should be ‘...administered by industry knowledgeable people’ (Newbery, 2003, p. 9). Similar considerations apply with at least as much force to scoping and analysing the sector through designing survey instruments, maxi- mizing response rates and interpreting the results.

References Allnutt, D. (2004). Review of tourism—Executive summary. Statistics tourism statistics improvement initiative. London: Department for Culture, Media and Support. British Design Initiative & Design Council (2002). The British design industry valuation survey 2002. London: British Design Initiative & Design Council. British Design Initiative & Design Council (2003). The British design industry valuation survey 2003. London: British Design Initiative & Design Council. Cabinet Office (1999). Modernising government (Cm4310). London: The Stationery Office. Cabinet Office Performance and Innovation Unit (2000). Adding it up. Improving analysis and modelling in central government. London: Performance and Innovation Unit. Campbell, D. T. (1974) Evolutionary epistemology. In P. Schilpp (Ed.), The philosophy of Karl Popper. La Salle: Open Court. Casey, B., Dunlop, R., & Selwood, S. (1996). Culture a commodity? The economics of the arts and built heritage in the UK. London: Policy Studies Institute. Cheshire, S. (1998). British fashion designer report. London: Department of Trade and Industry. Coates, C. (1997) Designer fact file: A guide to setting up a ‘Designer Fashion’ business. London: Department of Trade and Industry. Cochrane, A. L. (1972) Effectiveness and efficiency: Random reflections on health services. London: Nuffield Hospitals Trust. Creigh-Tyte, A., & Thomas, B. (2001) Employment. In S. Selwood (Ed.), The UK cultural sector profile and policy issues (Chapter 23, pp. 250–279). London: Policy Studies Institute. Creigh-Tyte, A., Creigh-Tyte, S., & Thomas, B. (2003). Employment in the cultural sector: The United Kingdom in the late 1990s. International Journal of Arts Management, 5(2), pp. 29–43. Creigh-Tyte, S., & Mundy, G. (2003). A research strategy for DCMS 2003–2005/06 (Tech. Paper No. 3). London: Department for Culture, Media and Sport. Creigh-Tyte, S., Gillham, J., & and Mundy, G. (2004). Evidence-based cultural policy in the UK. Paper presented at the Fourth Campbell Collaboration Colloquium, Washington, DC, February. Davies, H. T. O., Nutley, S. M. & Smith, P. C. (2000). What works? Evidence-based policy and practice in public services. Bristol: The Policy Press. Department for Culture, Media and Sport (2004a). Tomorrow’s tourism today. A prospectus for tourism development in five key areas. London: Department for Culture, Media and Sport. Department for Culture, Media and Sport (2004b). Creative Industries Economic Estimates. London: Department for Culture, Media and Sport. Department for Culture, Media and Sport (2004c). DCMS evidence toolkit—DET (formerly, the Regional Cultural Data Framework). London: Department for Culture, Media and Sport. 182 A. Creigh-Tyte Department for Culture, Media and Sport & Creative Industries Task Force (1998). Creative industries 1998: Mapping documents. London: Department for Culture, Media and Sport. Department for Culture, Media and Sport & the English Regional Cultural Consortia (2002a). Regional Cultural Data Framework: A user’s guide for researchers and policy makers. London, Department for Culture, Media and Sport. Department for Culture, Media and Sport & the English Regional Cultural Consortia (2002b). Regional Cultural Data Framework. Final technical report. London: Department for Culture, Media and Sport. Department for Culture, Media and Sport & Ministerial Creative Industries Strategy Group (2001). Creative industries Mapping Document 2001. London: Department for Culture, Media and Sport. Design Council (2004). Design in Britain 2003–2004. London: Design Council. Employment Department Group Office of Population Censuses and Surveys (1990). Standard Occupational Classification: Volume 1, 2 and 3. London: HMSO. Hanberger, A. (2001). What is the policy problem? Methodological challenges in policy evaluation. Evaluation, 7(1), pp. 45–62. HM Treasury (2003). The green book. Appraisal and evaluation in central government. London: The Stationery Office. Johnson, P., & Thomas, B. (2001). Assessing the economic impact of the arts. In S. Selwood (Ed.), The UK cultural sector profile and policy issues (Chapter 20, pp. 202–216). London: Policy Studies Institute. Jones, C., Munday, M., Bryan, J., Roberts, A., McNicoll, I. & McLellan, D. (2004). UK tourism satellite account—First steps project. London: Department for Culture, Media and Sport. Jones, G. (2000). The development of the annual business inquiry. Economic Trends, November, pp. 49–57. Kurt Salmon Associates (1991) British designer industry survey. Manchester: Kurt Salmon Associates. Labour Party (1997). Create the future. A strategy for cultural policy, arts and the creative economy. London: Labour Party. Mahajan, S. (2004). Creative sector, 1992–2002. Extract taken from United Kingdom input–output analyses, 2004 Edition. London: Office for National Statistics website articles database. Newbery, M. (2003). A study of the UK ‘Designer Fashion’ sector. Findings and recommendations. London: Malcolm Newbery Consulting Company. Office for National Statistics (1997) Indexes to the UK Standard Industrial Classification of Economic Activities 1992: Incorporating new sub classes created since the 1992 edition. London: Stationery Office. Office for National Statistics (2000). Standard Occupational Classifications 2000, volume 1. London: Stationery Office. Office for National Statistics (2002). UK Standard Industrial Classification of economic activities 2003. London: Stationery Office. Office of Science and Technology (2004). Science review of the DCMS. London: Department of Trade and Industry. Owen, N., & Cannon Jones, A. (2003). A comparative study for the British and Italian textile and clothing industries (DTI Economics Paper No. 2). London: Department of Trade and Industry. Power, D., & Scott, A. J. (2004). Cultural industries and the production of culture. Abingdon: Routledge. Pratt, A. (1997). The cultural industries production system: A case study of employment change in Britain, 1984–1991. Environment and Planning A, 29(11), 1953–1974. Pratt, A. (2004). The cultural economy. A call for spatialized ‘production of culture’ perspectives. International Journal of Cultural Studies, 7(1), 117–128. Rantisi, N. (2004). The designer in the city and the city in the designer. In D. Power & A. J. Scott (Eds.), Cultural industries and the production of culture (Chapter 6). Abingdon: Routledge. Cultural Trends 183 Roodhouse, S. (2003). Essential facts: The name of ‘Designer Fashion’ and its markets. Bolton: Bolton Institute. Scott, A. (2000). The cultural economy of cities. London: Sage. Textile and Clothing Strategy Group (2000). A national strategy for the UK textile and clothing industry. London: Department of Trade and Industry. Vervaeke, M., & Le Febvre, B. (2002). Design trades and inter-firm relationships in the Nord-Pas de Calais textile industry. Regional Studies, 36(6), 661–673. Young, K., Ashby, D., Boaz, A., & Grayson, L. (2002). Social science and the evidence-based policy movement. Social Policy and Society, 1(3), 215–224.